Converting from the Game Gear to the Master System | Hacker News Reader

Converting from the Game Gear to the Master System

nicole.express

117 pointsstevekemp18 comments

The Master System was very very similar to the MSX computer. AFAIK, it uses the same sound chip, the video modes are a subset of the other and the only real difference between both systems was memory mapping and access to peripherals.

My hypothesis is that SEGA wanted to compete against the 8bit famicom launching something that allowed easy porting of MSX games, so the console could quickly gain a lot of good games. I don't know why that didn't materialized. It probably has to do with exclusivity contracts between Nintendo and publishers and also because the Master System arrived late to the game. This is a shame because it was a considerably powerful compared to the famicom[1][2].

SEGA killed it quickly in the US and Japan and successfully competed on the next generation with the Genesis/Mega Drive against the Super NES. The Master System had some success in Europe and in Brazil it lasted at least until early 2000's. In late 90's it got a very advanced port, considering it was an 8-bit console, of Street Fighter 2 made by TecToy released exclusively in Brazil which is very a valuable item among collectors.

[1] https://www.copetti.org/images/consoles/mastersystem/diagram...

[2] https://www.copetti.org/images/consoles/nes/diagram.f88808f2...

The SG-1000, Sega's first console, uses the same chip combo as MSX (the Sega Mark III / Master System a couple of years later adds the extra video modes), but it was released a few months before MSX. It was also released on the same day as the Famicom. So the timing doesn't really work quite right for easy MSX ports being a design consideration.

The ColecoVision also uses the same set of chips, and was released before, however...

Was the Master System significantly more affordable than the MSX? If not, I could see how Japanese consumers would find it a better deal to buy the more capable full system instead. Reminds me of how Commodore's attempt to make a game console that was just a more limited C64 was dead in the water.

Master system was JP¥16,800 / US$200 / GB£99 in 85-87. https://en.wikipedia.org/wiki/Master_System

MSX was a standard and not a specific machine. Though, most machines probably would have been more money. You can check the prices of specific models and when they were released here, like the "simple" 1985 Toshiba HX-20 which sold for JP¥43,800. https://www.msx.org/wiki/Toshiba_HX-30

    This is a shame because it was a considerably 
    powerful compared to the famicom

This is 100% true: the base Master System is more powerful than the base NES in terms of RAM, and onscreen colors, ability to do split screen / multilayered scrolling, and a few other key metrics.

Though, the slightly fuller picture is that the Famicom's weirdly unique architecture made it able to be "upgraded" by hardware on the game carts themselves.

So in practice, many flagship NES titles quickly surpassed those on the Master System by various technical measures.

Nintendo also had something I think is a great idea when developing the NES: the video hardware had its own bus to the cartridge. This probably saved many cycles keeping the processor busy with other things instead of transferring content to the video RAM. It also allows better use of the system's very limited RAM and video RAM.

I don't know if there was any patent related to this, and I can't understand why SEGA didn't copy it. Maybe it could make the console more expensive since SEGA didn't produce their own chips and maybe they were really expecting to make ports easier.

I still think a bus connecting the video hardware to the cartridge could make the Master System way more powerful.

God, yeah. It's very difficult for me to look at retro game hardware design decisions without obsessing over "what if?" scenarios.

My favorite (and by that I mean, "most agonizing") what-if's concern the Genesis/Megadrive.

I don't have handy links, but:

- There was some sort of hardware fail that prevented clear sample playback. The sound chip could play back crystal clear samples but in reality, this was impossible to achieve thanks to some issue (DMA? bus? something?) and in reality, sampled sounds on the Genesis sounded like butt

- The Genesis' limited color palette and number of onscreen colors relative to the SNES and even the TG16 was kind of painful. I remember reading that one reason for this limitation was the need to include backwards compatibility with the SMS in the Genesis' video chip -- a feature few if any cared about or used. Arrrrgh.

A Genesis with color to rival the SNES/TG16 would have been such a fearsome beast.

> Nintendo also had something I think is a great idea when developing the NES: the video hardware had its own bus to the cartridge. This probably saved many cycles keeping the processor busy with other things instead of transferring content to the video RAM. It also allows better use of the system's very limited RAM and video RAM.

Nintendo did this because they had to, not because it was an ingenious innovation. Sega didn’t do it in the SMS/GG for the same reason Nintendo didn’t do it in the GB or GBA; because those architectures were much more efficient at ferrying data due to their high cycle counts.

The Z80, for the most part, at 4MHz is about equivalent to a 6502 at 1MHz due to architectural differences / differing design decisions. Except in strictly loop-based copies, since it can move more data in the same sliver of time, especially if you utilize ported memory.

> Sega didn’t do it in the SMS/GG for the same reason Nintendo didn’t do it in the GB or GBA; because those architectures were much more efficient at ferrying data due to their high cycle counts.

I'm still unconvinced. I wonder what could be done with that spare cycles and what the video hardware could do with a dedicated bus with a full address space all for itself.

There was nothing special about NES’s architecture. It was literally just a cart that, when plugged in, was directly wired to a subset of address lines. Utilizing the CHR/PRG separation allowed for some interesting mapper designs and eased graphics loading, sure. But the Master System was as capable of mapping cart based logic. The Master System just didn’t need it during its peak sales period (due to the aforementioned more capable base design) nor did it last long enough for developers to focus on developing anything special (Nintendo and a few major companies make up the developers of the majority of “official” mappers, Sega didn’t have many big names at the time nor did the support the system long enough to need it).

That being said, there totally were a few mappers developed for the console [1].

1 - https://www.smspower.org/Development/Mappers

> multilayered scrolling

I'm not exactly sure what you mean but both the NES and the SMS have one background layer and one sprite layer.

Yeah, I didn't use the correct term, really.

Most people would associate "multilayered scrolling" with multiple fully independent background layers like the Genesis had.

What the Master System did have was the necessary bits (interrupts? scan line counters? etc?) to alter the scroll registers during hblank intervals to do scroll various bits of a single background layer at different speeds.

Commercial example, "TransBot" a tiny budget card based game - https://youtu.be/k2U-SnDCJ4M?t=878

Amazing modern homebrew demo example - https://www.youtube.com/watch?v=SMjduDpUeaI

NES games that did this typically needed one of the mapper chips that provided the necessary counters/interrupts/idunnowhat. But the SMS can handle this with ease, no extra hardware required.

It's sort of similar to the MSX2, much more than it is to the MSX1.

I really enjoyed this author's writing style, it's rare that I laugh out loud at a technical article.

Absolutely agree. Balancing the extremely technical side of things with humor and readability is such a rare feat.

The author, Nicole, has consistently been one of the shining stars of this sort of technoarcheology.

She's writing games for these consoles, too, so it's not just dry archeology. It's not dead technology to her.

The Game Gear actually had an accessory to plug MS games into it. It literally just mapped the circuits to the smaller form factor.

I leaned on it heavily in the late 90s buying MS games from op shops.

https://segaretro.org/Master_Gear_Converter

I guess this makes sense from a business perspective. Why build something brand new when you can reuse.

I love the technical breakdown.

I've seen some interesting NES/Famicom conversions to PC Engine / TG16 in ROM collections. Apparently there are a lot more of them - here's a thread about how they were done: https://www.pcenginefans.com/forum/index.php?topic=23408.15

Basic idea is that both consoles use 6502-family CPUs, and vaguely similar tile + sprite graphics hardware, so the original machine code can mostly be run while adapting the specific register programming of one machine's graphics/sound hardware to the other - it helps that the target machine is much more capable.

Reply on news.ycombinator.com